DeepSLAM: A Robust Monocular SLAM System With Unsupervised Deep Learning

In this article, we propose DeepSLAM, a novel unsupervised deep learning based visual simultaneous localization and mapping (SLAM) system. The DeepSLAM training is fully unsupervised since it only requires stereo imagery instead of annotating ground-truth poses. Its testing takes a monocular image s...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	IEEE transactions on industrial electronics (1982) Jg. 68; H. 4; S. 3577 - 3587
Hauptverfasser:	Li, Ruihao, Wang, Sen, Gu, Dongbing
Format:	Journal Article
Sprache:	Englisch
Veröffentlicht:	New York IEEE 01.04.2021 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:	Artificial neural networks Coders Computer architecture Deep learning Depth estimation Imagery machine learning Optimization Outliers (statistics) Performance evaluation Pose estimation recurrent convolutional neural network (RCNN) Robustness Simultaneous localization and mapping simultaneous localization and mapping (SLAM) Three-dimensional displays Tracking Training unsupervised deep learning (DL) Visualization
ISSN:	0278-0046, 1557-9948
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Abstract	In this article, we propose DeepSLAM, a novel unsupervised deep learning based visual simultaneous localization and mapping (SLAM) system. The DeepSLAM training is fully unsupervised since it only requires stereo imagery instead of annotating ground-truth poses. Its testing takes a monocular image sequence as the input. Therefore, it is a monocular SLAM paradigm. DeepSLAM consists of several essential components, including Mapping-Net, Tracking-Net, Loop-Net, and a graph optimization unit. Specifically, the Mapping-Net is an encoder and decoder architecture for describing the 3-D structure of environment, whereas the Tracking-Net is a recurrent convolutional neural network architecture for capturing the camera motion. The Loop-Net is a pretrained binary classifier for detecting loop closures. DeepSLAM can simultaneously generate pose estimate, depth map, and outlier rejection mask. In this article, we evaluate its performance on various datasets, and find that DeepSLAM achieves good performance in terms of pose estimation accuracy, and is robust in some challenging scenes.
AbstractList	In this article, we propose DeepSLAM, a novel unsupervised deep learning based visual simultaneous localization and mapping (SLAM) system. The DeepSLAM training is fully unsupervised since it only requires stereo imagery instead of annotating ground-truth poses. Its testing takes a monocular image sequence as the input. Therefore, it is a monocular SLAM paradigm. DeepSLAM consists of several essential components, including Mapping-Net, Tracking-Net, Loop-Net, and a graph optimization unit. Specifically, the Mapping-Net is an encoder and decoder architecture for describing the 3-D structure of environment, whereas the Tracking-Net is a recurrent convolutional neural network architecture for capturing the camera motion. The Loop-Net is a pretrained binary classifier for detecting loop closures. DeepSLAM can simultaneously generate pose estimate, depth map, and outlier rejection mask. In this article, we evaluate its performance on various datasets, and find that DeepSLAM achieves good performance in terms of pose estimation accuracy, and is robust in some challenging scenes.
Author	Li, Ruihao Gu, Dongbing Wang, Sen
Author_xml	– sequence: 1 givenname: Ruihao orcidid: 0000-0002-9839-1489 surname: Li fullname: Li, Ruihao email: liruihao2008@gmail.com organization: Artificial Intelligence Research Center, National Innovation Institute of Defense Technology, Beijing, China – sequence: 2 givenname: Sen orcidid: 0000-0003-1537-8834 surname: Wang fullname: Wang, Sen email: s.wang@hw.ac.uk organization: Edinburgh Centre for Robotics, Heriot-Watt University, Edinburgh, U.K – sequence: 3 givenname: Dongbing orcidid: 0000-0002-0986-2921 surname: Gu fullname: Gu, Dongbing email: dgu@essex.ac.uk organization: School of Computer Science and Electronic Engineering, University of Essex, Colchester, U.K
BookMark	eNp9kMFLwzAUxoNMcJveBS8Bz50vado03sacbtAhuA2PpU1ftGNra9IK--9t2fDgwdM7vO_3ffAbkUFZlUjILYMJY6AeNsv5hAOHCVcRBxVekCELAukpJaIBGQKXkQcgwisycm4HwETAgiFZPCHW63i6eqRT-lZlrWvoqior3e5TS_sHXR9dgwf6XjSfdFu6tkb7XTjMaY_SGFNbFuXHNbk06d7hzfmOyfZ5vpktvPj1ZTmbxp7mijWeDLphY5QOWZ5ynfvopxwwENoIrYxAE-UgZWZYJlSOOssyH1kWmYyj5hj6Y3J_6q1t9dWia5Jd1dqym0y4kBD6vpK8S4WnlLaVcxZNoosmbYqqbGxa7BMGSW8t6awlvbXkbK0D4Q9Y2-KQ2uN_yN0JKRDxN65ASCbB_wHDU3lD
CODEN	ITIED6
CitedBy_id	crossref_primary_10_3390_aerospace11090768 crossref_primary_10_3390_rs15041156 crossref_primary_10_1002_rob_22454 crossref_primary_10_1145_3580854 crossref_primary_10_3390_robotics11010024 crossref_primary_10_1109_TITS_2022_3153815 crossref_primary_10_1016_j_patcog_2024_110573 crossref_primary_10_1109_TITS_2024_3462596 crossref_primary_10_1088_1361_6501_adfe04 crossref_primary_10_1109_TETCI_2024_3360329 crossref_primary_10_1109_JSEN_2024_3355907 crossref_primary_10_3390_electronics13163332 crossref_primary_10_1109_TIE_2022_3203761 crossref_primary_10_3390_app14177767 crossref_primary_10_3390_s25051447 crossref_primary_10_1109_TITS_2023_3284228 crossref_primary_10_1109_ACCESS_2023_3249661 crossref_primary_10_1016_j_pnucene_2025_105821 crossref_primary_10_1109_ACCESS_2023_3251733 crossref_primary_10_1016_j_procs_2024_11_035 crossref_primary_10_3390_s25165029 crossref_primary_10_1109_TIM_2025_3602581 crossref_primary_10_1177_01423312251344676 crossref_primary_10_32604_cmes_2022_019430 crossref_primary_10_3390_rs14133010 crossref_primary_10_1016_j_engappai_2024_108466 crossref_primary_10_1007_s10791_025_09607_0 crossref_primary_10_1016_j_displa_2024_102806 crossref_primary_10_1109_TITS_2021_3127646 crossref_primary_10_1109_ACCESS_2022_3144845 crossref_primary_10_1080_19479832_2024_2382737 crossref_primary_10_1007_s10846_022_01582_8 crossref_primary_10_1007_s12559_022_10010_w crossref_primary_10_1109_TVT_2024_3492388 crossref_primary_10_1088_1361_6501_ad9347 crossref_primary_10_1109_TIE_2022_3222696 crossref_primary_10_1109_TASE_2023_3330704 crossref_primary_10_1109_TVT_2023_3241634 crossref_primary_10_1109_TRO_2021_3120036 crossref_primary_10_1109_TCSVT_2025_3526645 crossref_primary_10_1016_j_robot_2025_105081 crossref_primary_10_1088_1361_6501_adcce2 crossref_primary_10_1109_ACCESS_2022_3199689 crossref_primary_10_1109_ACCESS_2023_3282637 crossref_primary_10_1109_ACCESS_2025_3589161 crossref_primary_10_1109_JSEN_2023_3324559 crossref_primary_10_1109_TIE_2022_3176304 crossref_primary_10_1109_LRA_2024_3440851 crossref_primary_10_1109_LRA_2024_3352357 crossref_primary_10_1109_TIE_2023_3327342 crossref_primary_10_3390_app14219748 crossref_primary_10_1088_1361_6501_ad824b crossref_primary_10_1016_j_jestch_2025_101977 crossref_primary_10_1109_TIV_2022_3173662 crossref_primary_10_1111_mice_13353 crossref_primary_10_1109_TIE_2022_3219119 crossref_primary_10_3390_rs14236033 crossref_primary_10_3390_s23042113 crossref_primary_10_1186_s13634_021_00795_7 crossref_primary_10_1016_j_jmapro_2025_07_063 crossref_primary_10_1007_s10489_024_05776_5 crossref_primary_10_3390_rs15215141 crossref_primary_10_3390_s24185925 crossref_primary_10_1007_s40815_024_01930_w crossref_primary_10_1016_j_neucom_2023_02_014 crossref_primary_10_3390_sym17040508 crossref_primary_10_1364_AO_465168 crossref_primary_10_3390_s21144735 crossref_primary_10_1007_s10922_024_09809_9
Cites_doi	10.1109/TASE.2017.2664920 10.1177/0278364917734298 10.1109/CVPR.2017.596 10.1109/TCI.2016.2644865 10.1177/0278364916679498 10.1109/ICCV.2017.75 10.1109/ICRA.2019.8793527 10.1109/ICRA.2017.7989236 10.1007/978-3-030-28619-4_38 10.1109/CVPR.2017.699 10.1007/978-3-319-10605-2_54 10.1109/ICME.2018.8486548 10.1109/TRO.2015.2463671 10.1109/CVPR.2017.700 10.1109/TPAMI.2007.1049 10.1007/978-3-319-46484-8_45 10.1007/978-3-319-46493-0_51 10.1109/ICCV.2015.336 10.1109/TIE.2018.2854557 10.1109/IVS.2011.5940405 10.1007/s10514-015-9516-2 10.1109/CVPR.2017.695 10.1109/ICRA.2018.8461251 10.1109/ICCV.2011.6126513 10.1109/TIP.2003.819861 10.1109/ICRA.2014.6906953 10.1007/978-3-319-70353-4_57 10.1109/LRA.2015.2505717 10.1109/TPAMI.2017.2658577 10.1109/IROS.2017.8202236 10.1109/TRO.2012.2197158 10.1109/ICInfA.2018.8812582 10.1109/TIE.2018.2826471 10.1109/CVPR.2012.6248074 10.1109/ICRA.2016.7487679 10.1109/ISMAR.2007.4538852
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
DBID	97E RIA RIE AAYXX CITATION 7SP 8FD L7M
DOI	10.1109/TIE.2020.2982096
DatabaseName	IEEE Xplore (IEEE) IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Electronics & Communications Abstracts Technology Research Database Advanced Technologies Database with Aerospace
DatabaseTitle	CrossRef Technology Research Database Advanced Technologies Database with Aerospace Electronics & Communications Abstracts
DatabaseTitleList	Technology Research Database
Database_xml	– sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	1557-9948
EndPage	3587
ExternalDocumentID	10_1109_TIE_2020_2982096 9047170
Genre	orig-research
GrantInformation_xml	– fundername: DeepField Project grantid: 857339 – fundername: Engineering and Physical Sciences Research Council funderid: 10.13039/501100000266 – fundername: National Natural Science Foundation of China grantid: 61903377 funderid: 10.13039/501100001809 – fundername: Artificial Intelligence Offshore Robotics for Certification of Assets grantid: EP/R026173/1 – fundername: EU H2020 Program under EUMarineRobots Project grantid: 731103
GroupedDBID	-~X .DC 0R~ 29I 4.4 5GY 5VS 6IK 97E 9M8 AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACGFS ACIWK ACKIV ACNCT AENEX AETIX AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD HZ~ H~9 IBMZZ ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 MS~ O9- OCL P2P RIA RIE RNS TAE TN5 TWZ VH1 VJK AAYXX CITATION 7SP 8FD L7M
ID	FETCH-LOGICAL-c291t-75145ff9c61da2cd3e3a20e54cf4c9f4ef8d077bf1b49decbbb3e1b8fb2ec2e63
IEDL.DBID	RIE
ISICitedReferencesCount	96
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000599525100077&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	0278-0046
IngestDate	Mon Jun 30 10:14:08 EDT 2025 Tue Nov 18 21:39:50 EST 2025 Sat Nov 29 01:31:47 EST 2025 Wed Aug 27 02:33:09 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	4
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c291t-75145ff9c61da2cd3e3a20e54cf4c9f4ef8d077bf1b49decbbb3e1b8fb2ec2e63
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0003-1537-8834 0000-0002-9839-1489 0000-0002-0986-2921
PQID	2470633972
PQPubID	85464
PageCount	11
ParticipantIDs	proquest_journals_2470633972 crossref_primary_10_1109_TIE_2020_2982096 ieee_primary_9047170 crossref_citationtrail_10_1109_TIE_2020_2982096
PublicationCentury	2000
PublicationDate	2021-04-01
PublicationDateYYYYMMDD	2021-04-01
PublicationDate_xml	– month: 04 year: 2021 text: 2021-04-01 day: 01
PublicationDecade	2020
PublicationPlace	New York
PublicationPlace_xml	– name: New York
PublicationTitle	IEEE transactions on industrial electronics (1982)
PublicationTitleAbbrev	TIE
PublicationYear	2021
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref35 ref13 ref34 ref12 ref37 ref14 mohanty (ref17) 2016 ref31 ref30 ref33 ref11 ref2 detone (ref23) 2017 ref1 szegedy (ref43) 0 ref38 ref16 ref19 ref18 simonyan (ref39) 2015 eigen (ref10) 0 ref24 ref45 ref26 ref25 ref20 ref42 ref41 ref22 ref44 ref21 kümmerle (ref36) 0 clark (ref15) 0 ref27 ref29 ref8 jaderberg (ref28) 0 ref7 ref9 vijayanarasimhan (ref32) 0 ref4 ref3 ref6 ref5 ref40
References_xml	– ident: ref12 doi: 10.1109/TASE.2017.2664920 – ident: ref19 doi: 10.1177/0278364917734298 – ident: ref22 doi: 10.1109/CVPR.2017.596 – ident: ref41 doi: 10.1109/TCI.2016.2644865 – year: 0 ident: ref32 article-title: SfM-Net: Learning of structure and motion from video publication-title: arXiv 1704 07804 – ident: ref45 doi: 10.1177/0278364916679498 – ident: ref13 doi: 10.1109/ICCV.2017.75 – ident: ref27 doi: 10.1109/ICRA.2019.8793527 – ident: ref18 doi: 10.1109/ICRA.2017.7989236 – year: 2016 ident: ref17 article-title: DeepVO: A deep learning approach for monocular visual odometry – ident: ref21 doi: 10.1007/978-3-030-28619-4_38 – start-page: 2366 year: 0 ident: ref10 article-title: Depth map prediction from a single image using a multi-scale deep network publication-title: Proc Adv Neural Inf Process Syst – start-page: 4278 year: 0 ident: ref43 article-title: Inception-v4, Inception-ResNet and the impact of residual connections on learning publication-title: Proc AAAI Conf Artif Intell – ident: ref30 doi: 10.1109/CVPR.2017.699 – start-page: 2017 year: 0 ident: ref28 article-title: Spatial transformer networks publication-title: Proc Adv Neural Inf Process Syst – ident: ref7 doi: 10.1007/978-3-319-10605-2_54 – ident: ref26 doi: 10.1109/ICME.2018.8486548 – ident: ref5 doi: 10.1109/TRO.2015.2463671 – ident: ref31 doi: 10.1109/CVPR.2017.700 – ident: ref1 doi: 10.1109/TPAMI.2007.1049 – ident: ref11 doi: 10.1007/978-3-319-46484-8_45 – ident: ref29 doi: 10.1007/978-3-319-46493-0_51 – year: 2017 ident: ref23 article-title: Toward geometric deep SLAM – ident: ref9 doi: 10.1109/ICCV.2015.336 – ident: ref3 doi: 10.1109/TIE.2018.2854557 – ident: ref42 doi: 10.1109/IVS.2011.5940405 – ident: ref37 doi: 10.1007/s10514-015-9516-2 – ident: ref24 doi: 10.1109/CVPR.2017.695 – start-page: 1 year: 2015 ident: ref39 article-title: Very deep convolutional networks for large-scale image recognition publication-title: Int Conf Learn Representations – ident: ref33 doi: 10.1109/ICRA.2018.8461251 – start-page: 6856 year: 0 ident: ref15 article-title: VidLoc: 6-DoF video-clip relocalization publication-title: Proc Conf Comput Vis and Pattern Recog – ident: ref6 doi: 10.1109/ICCV.2011.6126513 – ident: ref40 doi: 10.1109/TIP.2003.819861 – ident: ref35 doi: 10.1109/ICRA.2014.6906953 – ident: ref20 doi: 10.1007/978-3-319-70353-4_57 – ident: ref16 doi: 10.1109/LRA.2015.2505717 – ident: ref8 doi: 10.1109/TPAMI.2017.2658577 – ident: ref38 doi: 10.1109/IROS.2017.8202236 – ident: ref34 doi: 10.1109/TRO.2012.2197158 – ident: ref25 doi: 10.1109/ICInfA.2018.8812582 – ident: ref4 doi: 10.1109/TIE.2018.2826471 – ident: ref44 doi: 10.1109/CVPR.2012.6248074 – ident: ref14 doi: 10.1109/ICRA.2016.7487679 – ident: ref2 doi: 10.1109/ISMAR.2007.4538852 – start-page: 3607 year: 0 ident: ref36 article-title: g2o: A general framework for graph optimization publication-title: Proc IEEE Int Conf Robot Autom
SSID	ssj0014515
Score	2.6462696
Snippet	In this article, we propose DeepSLAM, a novel unsupervised deep learning based visual simultaneous localization and mapping (SLAM) system. The DeepSLAM...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	3577
SubjectTerms	Artificial neural networks Coders Computer architecture Deep learning Depth estimation Imagery machine learning Optimization Outliers (statistics) Performance evaluation Pose estimation recurrent convolutional neural network (RCNN) Robustness Simultaneous localization and mapping simultaneous localization and mapping (SLAM) Three-dimensional displays Tracking Training unsupervised deep learning (DL) Visualization
Title	DeepSLAM: A Robust Monocular SLAM System With Unsupervised Deep Learning
URI	https://ieeexplore.ieee.org/document/9047170 https://www.proquest.com/docview/2470633972
Volume	68
WOSCitedRecordID	wos000599525100077&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1557-9948 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014515 issn: 0278-0046 databaseCode: RIE dateStart: 19820101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3NS8MwFA9zeNCDX1OcTsnBi2C3Ns2axttQx4Q5RDfcrTTJqw5kG2vr32-SdmWgCN4KzYPye3lfzcvvIXQFHpNhoBLHdzlzqBdyRwhJHK5zeyZjro3ManrIRqNwOuXPNXRT3YUBANt8Bm3zaM_y1ULm5ldZh7valTJdoG8xFhR3taoTA9otphUQwxiri771kaTLO-PHB10IErdNuI53hp5_IwTZmSo_HLGNLv39_33XAdors0jcK9R-iGowP0K7G9yCDTS4B1i-DntPt7iHXxYiTzOsDXhh-06xeYELtnL8Nss-8GSe5kvjN1JQ2Ijiknn1_RhN-g_ju4FTjk1wJOFe5jCdA3WThMvAUzGRygc_Ji50qUyo5AmFJFQuYyLxBOUKpBDCB0-EiSAgCQT-CarPF3M4RRi0xQJw6mqtUQ5eDL6SLKZEMQ8Cwpqos0YykiWnuBlt8RnZ2sLlkcY-MthHJfZNdF1JLAs-jT_WNgzW1boS5iZqrZUVlQaXRoQynWzp5Iqc_S51jnaIaUexTTctVM9WOVygbfmVzdLVpd1L30PsxbM
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1bS8MwFD6ICuqDtynOax58EaxL02xpfBtemDiH6MS9lSY5VUG2YTt_v0naDUERfCs0B8p3cm7NyXcAjjEUOm6ZLIioFAEPYxkopVkgbW4vdCqtkXlNd0WvFw8G8n4OTmd3YRDRN5_hmXv0Z_lmpCfuV1lDUutKhS3QF5qcM1re1pqdGfBmOa-AOc5YW_ZNDyWpbPRvrmwpyOgZkzbiOYL-b0HIT1X54Yp9fLle-9-XrcNqlUeSdqn4DZjD4SasfGMXrEHnEnH82G3fnZM2eRipSV4Qa8Ij33lK3AtS8pWT57filTwN88nYeY4cDXGipOJefdmCp-ur_kUnqAYnBJrJsAiEzYKaWSZ1KzQp0ybCKGUUm1xnXMuMYxYbKoTKQsWlQa2UijBUcaYYaoataBvmh6Mh7gBBa7OIklOrNy4xTDEyWqScGRFii4k6NKZIJrpiFXfDLd4TX11QmVjsE4d9UmFfh5OZxLhk1Phjbc1hPVtXwVyH_amyksrk8oRxYdMtm16x3d-ljmCp07_rJt2b3u0eLDPXnOJbcPZhvviY4AEs6s_iLf849PvqC7f8yPo
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=DeepSLAM%3A+A+Robust+Monocular+SLAM+System+With+Unsupervised+Deep+Learning&rft.jtitle=IEEE+transactions+on+industrial+electronics+%281982%29&rft.au=Li%2C+Ruihao&rft.au=Wang%2C+Sen&rft.au=Gu%2C+Dongbing&rft.date=2021-04-01&rft.issn=0278-0046&rft.eissn=1557-9948&rft.volume=68&rft.issue=4&rft.spage=3577&rft.epage=3587&rft_id=info:doi/10.1109%2FTIE.2020.2982096&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TIE_2020_2982096
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0278-0046&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0278-0046&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0278-0046&client=summon